Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    PID controller algorithm

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     preservation
    0.38
     preserv
    0.38
     she
    0.37
     tanning
    0.37
    agli
    0.36
     حفظ
    0.36
     affordable
    0.36
     Fuss
    0.36
     Style
    0.35
     சூழ
    0.35
    POSITIVE LOGITS
     реки
    0.46
     ríos
    0.40
     retriever
    0.40
     njegove
    0.38
     हुँ
    0.38
     began
    0.37
     позиции
    0.36
    magazine
    0.36
    ウェ
    0.36
     බෙ
    0.36
    Act Density 0.001%

    No Known Activations